Overview
The PSI™ Video Compression engine enables Wellbeck Media Services to deliver high quality video material with a much lower band-width usage than any of our competitors. The bit rates for streaming can be varied according to the target device, the band-width capacity to be used and the quality required. The compression engine has a rich list of features that can flow to any media application that requires delivery of high quality video and audio:
- Delivers D1, DVD quality video at 1.1Mbs - 1/6th the bitrate of MPEG-2
- Delivers 3/4 D1, visual DVD quality video as low as 700Kbps
- Boasts one of the best signal to noise ratios of any codec on the market today
- Real-time encoding
- Region-based encoding that maintains image quality
- Allows embedding of popular encryption features and digital rights management systems (DRM)
Real-time encoding allows video material to be streamed as it is generated. The process is parameterised allowing both the video and audio elements to be streamed at bit rates which are suitable for the available band-width and the target device.
The video source would generally be fed directly into the hardware encoder and the output is then uploaded directly via an internet connection to the server from which the stream is to be effected.
Peak Signal to Noise Ratio Codec Comparison Chart
To assess a video codec's ability, the objective empirical test of peak signal to noise ratio (PSNR), defines the acid test for codec technology. The PSI_V™ codec has been benchmarked against our competitors and test results have shown the PSI_V™ codec is clearly superior in both PSNR values and visual quality in side by side tests.
PSNR is calculated in native YUY2 colour space. Any frames dropped (duplicated frames on the decode side) were ignored in the calculation. Hence codecs that drop frames such as Real, On2 and WM9 would have artificially 'enhanced' PSNR figures.
PSI_V TM Rate Reach Comparison
This rate reach chart shows the inherent limitations competitive compression technologies have in reaching the last mile
compared with PSI™ compressed video streams.
Resistance along the ADSL line grows as distance from the exchange source increases which results in a gradual
degradation of available band-width. Our compression
technology reduces the demand on that key resource.
Approach to compression
The key features of compactness, scalability, and priority have been attained as the main features of the PSI_V™ technology. PSI_V™ technology is an open and general-purpose architecture that effectively integrates these three features into a single platform for digital media coding.
Building on the mathematical foundation of video compression algorithms, the developers have perceived the fundamental issue of transform-based coding to be a problem of coding optimization, that is, how the transform coefficients are packed in the compressed bitstream.
Integrity requires that a minimized size of packed bit-stream be achieved subject to predefined quality measures (Peak Signal to Noise Ratio and Just-Noticeable Difference). The priority, scalability and integrity approach to video coding, applied to our video codec enabled us to achieve superior PSNR and JND results by not resorting to common 'optimization techniques' such as dropping frames during the encode and pre-/post-processing the video stream.
Scalability requires that the packed bit-stream be partially decodable such that it can be properly truncated during transmission or storage.
Priority prefers that visual objects or generic regions in the image be coded and packed at different levels of integrity, serving as basis for further explanation, manipulation, and retrieval.
The following two illustrations show how the regional encoding approach differs from the traditional based encoding approach:
The PSI_V™ codec treats what it sees more like the human eye than the traditional DCT based approach by applying selective compression ratios to varying parts of the scene. This enables the PSI_V™ codec to apply higher compression to less important aspects of a scene such as a static background element, and lower compression ratios to more visually important elements, such as the house in this example.
Coupled with integrity and scalability, varying compression within a given frame through priority differentiation helps the PSI_V™ codec achieve superior compression results.